The energy crisis and climate change led the global community towards renewable resources and wind energy, being clean and abundant, has a significant potential to generate power on a large scale. Though designing the layout of wind farm is crucial part to maximize energy production, the control settings of turbine give an advantage to optimize the control parameters based on the variations of wind while maximizing the energy production from a wind farm. But it is a challenging task to optimize the control settings of a turbine through model-based methods due to the nonlinear interactions of wake between the turbines. In contemporary times, reinforcement learning (RL) has been emerging as a promising area of research in the field of wind farm control. In this work, the authors aimed to use model-free deep deterministic policy gradient (DDPG) method-based reinforcement learning techniques to optimize the control settings of wind turbines to deflect the wakes and increase the power production. As a demonstration case study, yaw misalignment is considered as control parameter while maximizing the power production from a wind farm layout, which was obtained through robust layout optimization studies presented earlier.